From mboxrd@z Thu Jan  1 00:00:00 1970
From: Takashi Iwai <tiwai@suse.de>
Subject: Re: [patch] voluntary-preempt-2.6.9-rc1-bk4-Q5
Date: Tue, 31 Aug 2004 20:19:20 +0200
Sender: linux-kernel-owner@vger.kernel.org
Message-ID: <s5hllfvyxuf.wl@alsa2.suse.de>
References: <OF923A124A.1D8E364E-ON86256F01.0053F7B2-86256F01.0053F7D7@raytheon.com>
	<1093972819.5403.8.camel@krustophenia.net>
Mime-Version: 1.0 (generated by SEMI 1.14.5 - "Awara-Onsen")
Content-Type: text/plain; charset=US-ASCII
Return-path: <linux-kernel-owner+glk-linux-kernel=40m.gmane.org-S266291AbUHaSTi@vger.kernel.org>
In-Reply-To: <1093972819.5403.8.camel@krustophenia.net>
To: Lee Revell <rlrevell@joe-job.com>
Cc: Mark_H_Johnson@raytheon.com, Ingo Molnar <mingo@elte.hu>, "K.R. Foley" <kr@cybsft.com>, linux-kernel <linux-kernel@vger.kernel.org>, Felipe Alfaro Solana <lkml@felipe-alfaro.com>, Daniel Schmitt <pnambic@unu.nu>, alsa-devel <alsa-devel@lists.sourceforge.net>
List-Id: alsa-devel@alsa-project.org

At Tue, 31 Aug 2004 13:20:19 -0400,
Lee Revell wrote:
> 
> On Tue, 2004-08-31 at 11:17, Mark_H_Johnson@raytheon.com wrote:
> > >I will be running some additional tests
> > >- reducing preempt_max_latency
> > >- running with sortirq and hardirq_preemption=0
> > >to see if these uncover any further problems.
> > 
> > #1 - audio driver
> >  latency: 621 us, entries: 28 (28)
> >     -----------------
> >     | task: latencytest/11492, uid:0 nice:0 policy:1 rt_prio:99
> >     -----------------
> >  => started at: snd_ensoniq_playback1_prepare+0x74/0x180
> >  => ended at:   snd_ensoniq_playback1_prepare+0x11d/0x180
> > =======>
> > 00000001 0.000ms (+0.000ms): snd_ensoniq_playback1_prepare
> > (snd_pcm_do_prepare)
> > 00000001 0.014ms (+0.014ms): snd_es1371_dac1_rate
> > (snd_ensoniq_playback1_prepare)
> > 00000001 0.014ms (+0.000ms): snd_es1371_wait_src_ready
> > (snd_es1371_dac1_rate)
> > 00000001 0.562ms (+0.548ms): snd_es1371_src_read (snd_es1371_dac1_rate)
> > 00000001 0.562ms (+0.000ms): snd_es1371_wait_src_ready
> > (snd_es1371_src_read)
> > 00000001 0.578ms (+0.015ms): snd_es1371_wait_src_ready
> > (snd_es1371_src_read)
> > 00000001 0.585ms (+0.006ms): snd_es1371_src_write (snd_es1371_dac1_rate)
> > 00000001 0.585ms (+0.000ms): snd_es1371_wait_src_ready
> > (snd_es1371_src_write)
> > 00000001 0.601ms (+0.015ms): snd_es1371_src_write (snd_es1371_dac1_rate)
> > 00000001 0.601ms (+0.000ms): snd_es1371_wait_src_ready
> > (snd_es1371_src_write)
> > 00000001 0.602ms (+0.001ms): snd_es1371_wait_src_ready
> > (snd_es1371_dac1_rate)
> > 00000001 0.616ms (+0.013ms): smp_apic_timer_interrupt
> > (snd_ensoniq_playback1_prepare)
> > 
> > or
> > 
> >  latency: 663 us, entries: 41 (41)
> >     -----------------
> >     | task: latencytest/11492, uid:0 nice:0 policy:1 rt_prio:99
> >     -----------------
> >  => started at: snd_ensoniq_playback1_prepare+0x74/0x180
> >  => ended at:   snd_ensoniq_playback1_prepare+0x11d/0x180
> > =======>
> > 00000001 0.000ms (+0.000ms): snd_ensoniq_playback1_prepare
> > (snd_pcm_do_prepare)
> > 00000001 0.004ms (+0.004ms): snd_es1371_dac1_rate
> > (snd_ensoniq_playback1_prepare)
> > 00000001 0.005ms (+0.000ms): snd_es1371_wait_src_ready
> > (snd_es1371_dac1_rate)
> > 00000001 0.006ms (+0.001ms): snd_es1371_src_read (snd_es1371_dac1_rate)
> > 00000001 0.006ms (+0.000ms): snd_es1371_wait_src_ready
> > (snd_es1371_src_read)
> > 00000001 0.019ms (+0.012ms): snd_es1371_wait_src_ready
> > (snd_es1371_src_read)
> > 00000001 0.607ms (+0.588ms): snd_es1371_src_write (snd_es1371_dac1_rate)
> > 00000001 0.608ms (+0.000ms): snd_es1371_wait_src_ready
> > (snd_es1371_src_write)
> > 00000001 0.624ms (+0.016ms): snd_es1371_src_write (snd_es1371_dac1_rate)
> > 00000001 0.624ms (+0.000ms): snd_es1371_wait_src_ready
> > (snd_es1371_src_write)
> > 00000001 0.626ms (+0.001ms): snd_es1371_wait_src_ready
> > (snd_es1371_dac1_rate)
> > 00000001 0.639ms (+0.013ms): smp_apic_timer_interrupt
> > (snd_ensoniq_playback1_prepare)
> 
> Hmm, looks like the es1371 takes ~0.5 ms to set the DAC rate.  The ALSA
> team would probably be able to help.  Takashi, any ideas?

Does the attached patch fix this problem?


Takashi

--- linux/sound/pci/ens1370.c	25 Aug 2004 09:57:03 -0000	1.64
+++ linux/sound/pci/ens1370.c	31 Aug 2004 18:17:45 -0000
@@ -513,6 +513,7 @@
 		r = inl(ES_REG(ensoniq, 1371_SMPRATE));
 		if ((r & ES_1371_SRC_RAM_BUSY) == 0)
 			return r;
+		cond_resched();
 	}
 	snd_printk("wait source ready timeout 0x%lx [0x%x]\n", ES_REG(ensoniq, 1371_SMPRATE), r);
 	return 0;